Tuesday, September 18, 2018

More Rustbelt Woes for GOP, Senate Bright Spots for Democrats

Polls from Wisconsin, Arizona and Tennessee continue to show dramatic Democratic overperformance among white college graduate voters and Republican underperformance among white noncollege voters.
A new Marquette University poll shows Democrat Tony Evers continuing to lead Scott Walker in the Wisconsin governor's race. Evers' 5 point lead among likely voters is built on a 16 point lead among white college voters, while holding Walker to a mere 5 point lead among white noncollege voters (Trump carried them by 19 points in 2016). The poll also provide the breakdown by education and gender, which is quite interesting. Evers' has an amazing 26 point lead among white college grad women (62-36) and is actually carrying white noncollege women by 2 points (48-46). The latter finding underscores a point I and others have made that white noncollege women are a very vulnerable part of the Trump coalition and could undermine GOP fortunes across the country.
We see the same white college/noncollege pattern in some positive results for Democrats from CNN polls of likely voters in the Arizona and Tennessee Senate races. In Arizona, Krysten Sinema (D) is leading Martha McSally (R) by 7 points, with a healthy 11 point lead among white college graduates and a relatively modest 18 point deficit among white noncollege voters (Trump carried them by 27 points in 2016). In Tennessee, Phil Bredesen (D) is leading Marsha Blackburn (R) by 5 points, with a 5 point lead among white college grads and a 17 point deficit among white noncollege voters. Both figures are truly massive improvements over Trump's 2016 dominance, where he carried both white college and noncollege voters by huge margins (35 and 56 points, respectively).
Of course, even if Democrats pull out both these Senate races, that doesn't mean they will actually take the Senate--many, many incumbent Democrats in red states are in tight races. But these results are surely good news, both for what they tell us about these particular races and about evolving fractures in the Trump coalition.
About this website
JSONLINE.COM
Democrat Tony Evers saw a boost in support in a Tuesday poll that put him 5 points ahead of GOP Gov. Scott Walker a month after the two were tied.

Monday, September 17, 2018

Et Tu, Ohio?

Politico is out with a new poll of Ohio voters which underscores just how contested the Rustbelt has become in the Trump era. Far from being lost to the Democrats in the wake of the Trump election, Democrats are coming roaring back, riding a wave of discontent with the Orange One and his party.
In the new poll, Democrats lead the Congressional ballot by 7 points in the state, Democratic Senatorial incumbent Sherrod Brown is crushing challenger Jim Renacci and Democrat Richard Cordray is running about even with Mike DeWine in the governor's race
Clearly something good is going on for the Democrats in this state. Recall that Ohio swung sharply toward the GOP in 2016, supporting Trump by an 8 point margin. This was above all driven by a massive shift in his favor among white noncollege voters, who gave him a whopping 32 point margin. This poll and other data suggest that that margin is being whittled down considerably.
About this website
POLITICO.COM
In 2016, Ohio voters liked the president’s pitch on trade and jobs. That may be changing.

Saturday, September 15, 2018

House Update: Steady As She Goes

Not a lot new to report here. But a couple of interesting items.
* The latest CNN poll (+10 D on the generic) confirms a pattern I've seen a lot lately. The Democrats' lead among white college graduates is quite similar in magnitude to the Republicans' lead among white noncollege voters. The result is close to an even split among white voters overall. That's potentially catastrophic for the GOP.
* Harry Enten on the CNN site has an interesting piece examining the slight disjuncture between results implied by district-level House polls and overall generic results and various
"fundamental" indicators. Both imply the Democrats will take the House but the former implies smaller gains than the latter. Enten notes that the winning party on the overall popular vote tends to overperform their average on district-level polls, so there may be considerable upside for the Democrats relative to their district-level poll results.
Standard disclaimer: (a) None of this means the Democrats are certain to take the House. They aren't. (b) None of this means the Democrats can relax and not worry about voter mobilization. They can't.

Thursday, September 13, 2018

Yet More House Forecasting Models!

Larry Sabato's Crystal Ball just released a couple more academic House forecasting models. I'll spare you the grisly methodological details, but both models have the Democrats picking up 44 seats. Not bad, not bad at all.
I noticed something looking across the all the various Crystal Ball models and thinking about some other recent models/estimates of seat gain for the Democrats. Here's what I noticed:
average gain across all Crystal Ball academic models: 36 seats
538 model simulation average gain: 39 seats
average out party gain with unpopular president: 37 seats
Interesting convergence......
CRYSTALBALL.CENTERFORPOLITICS.ORG
On Tuesday, Nov. 6, about 90 million American voters (around 40% of the voting-eligible population, give or take) will elect all 435 members of the U.S. House…

The Blue Wave and the State Elections

It's not just about the House and the Senate. The outcomes of the various governors' and state legislative races will be absolutely crucial for our politics going. Here's an analysis of where we are on these races with one of my favorite analysts, Henry Olsen:
"Most discussions of the midterms focus on federal races for the House and the Senate. But nearly all states hold legislative elections this year, too, and 37 states will elect governors this fall. If the negative trends apparent at the federal level hit Republicans here, too, the repercussions could be felt in the party for years.
One might think that state and federal races would have different outcomes. After all, presidents and Congress have little to do with the issues that normally dominate state-level campaigns. Nevertheless, outcomes at each level tend to run in similar directions and with roughly equivalent dimensions.
Take the GOP “wave” election of 2010. Republicans not only picked up a massive 63 House and six Senate seats; they also gained six governorships and 680 state legislative seats. Democrats celebrated similar numbers four years earlier. In 2006, a Democratic landslide saw that party gain 31 House seats, six Senate seats, six governorships, and over 300 state legislative seats. Individual races might diverge from the national trend, but a wave will sweep all before it, no matter what level the race is at.
This bodes ill for Republicans running for governor this year. In part because many of the Republicans who won in 2010 are now term-limited out of office, Republicans must defend 11 governor’s chairs without the benefit of a previously elected incumbent. RealClearPolitics rates three of those seats as leaning to the Democrats with another six rated as toss-ups, including the large states of Florida, Georgia, and Ohio. GOP incumbents are also running in seven other states that by polling or by prior voting behavior raise caution flags. RealClear rates one of them, Illinois’ Bruce Rauner, as a likely loser, while incumbents in Wisconsin (Scott Walker) and Arizona (Doug Ducey) are considered toss-ups. Should Republicans lose the four races they trail in and only split the toss-ups, that would mean a net loss of eight governorships, the largest any party has lost in one election since the GOP picked up 11 in the historic 1994 wave.
Hundreds of state legislative seats are also in play. There’s lots of debate over when a seat should be considered vulnerable, but in the current environment most observers would probably agree that any Republican-held seat carried by either Hillary Clinton or Donald Trump with less than 50 percent of the vote qualifies for that distinction. A staggering 477 GOP-held statehouse seats fit these criteria. Another 202 Republican-held state Senate seats also fit the bill, though not all will be up for election because many states stagger senatorial terms. Losses of that magnitude would easily be the largest Republican losses in decades."
Wave on, Big Blue, wave on!
About this website
CITY-JOURNAL.ORG
Whatever happens nationally in November will likely be mirrored at the local level.

Wednesday, September 12, 2018

House Update: Looking Better and Better!

The 538 model now gives the Democrats an 83 percent (!) chance of taking the House, with an average seat gain of 39 seats (!!). Their rolling average on the House generic ballot has the Dems up by over 9 points.
Two recent polls by Quinnipiac and NPR/Marist have the Democrats up by 14 and 12 points, respectively in the generic House ballot. Fascinatingly, the Quinnipiac poll has the Democrats only losing white voters by 3 points, while the NPR/Marist poll has the Democrats actually carrying whites by 3 points. These are figures that should send a cold chill down the spines of GOP politicians. It's starting to look like white voters, in particular white noncollege voters, are not going to be enough to bail the Republicans out this time.

Sunday, September 9, 2018

Inside the Sweden Election Results

The Sweden election results are mostly in and here are the most salient points.
* The social democrats did a bit better than the polls indicated that they would. But their 28.4 percent performance was still historically bad and about 3 points down from the last election in 2014.
* The Sweden Democrats, the far-right populist party, which some feared would surge enough to become the country's second largest--or even largest--party, somewhat underperformed their pre-election polls. However, their 17.6 percent of the vote was still almost 5 points up from their 2014 performance, so they were the election's big winners.
* The overall center-left bloc (social democrats, greens, Left party) did very slightly better than the center-right bloc (Moderate party, Centre party, Liberal party, Christian Democrats). Current results show 40.6 percent of the vote and 144 parliamentary seats for the center-left bloc to 40.3 percent of the vote and 142 seats for the center-right bloc.
It is not yet clear what government will form based on these results. But, in theory, the other parties have committed to not including the Sweden Democrats in government, so whatever government forms will not include them.
The big question everyone is asking is: where did the Sweden Democrats come from? Who supports them and why? Since the Sweden Democrats are very hard-line on immigrants and immigration, as are similar European right populist parties, the dominant interpretation has tended to be that supporters of the Sweden Democrats are simply reacting in a racist and xenophobic way to the arrival of people who are different from them. Sweden is doing well economically so economic factors couldn't possibly have anything to do with it. (Sound familiar?)
A terrific paper published right before the election casts considerable doubt on this interpretation and provides a compelling alternative. This is one of the best papers I have seen on European right populism--I strongly recommend it.
The paper is Economic Losers and Political Winners: Sweden's Radical Right by Ernesto Dal Bo, Frederico Finan, Olle Folke,
Torsten Persson, and Johanna Rickne. Their basic approach and argument, including a plausible interaction between economic changes and immigration issues:
First and foremost, our study sheds new light on how economic shocks may explain the rise of the radical right. We start from the timing of the Sweden-Democrat rise: growing to enter parliament between 2006 and 2010, and continuing to become Sweden’s third largest party in 2014 (with a 12.9 percent vote share). This period pre-dated the 2015 refugee crisis, but coincided with two events that worsened the relative economic lot for large segments of the population. In 2006, a Center-Right coalition of parties took power and implemented a dramatic reform agenda of tax cuts and social-insurance austerity with the purpose to “make work pay”. Over a mere six years, these reforms triggered a dramatic increase in income inequality. With earned income tax credits, incomes continued to grow among “insiders” with stable employment, while spending cuts implied a stagnation of disposable incomes for “outsiders” with unstable or no jobs. The second key event is the 2008 financial crisis. The crisis increased the job insecurity for “vulnerable insiders”, segments of the population with stable employment, but with jobs at higher risk of replacement by automation and other forms of rationalization than “secure insiders”….
We find that the groups which faced a relative-income decline and higher job insecurity are over-represented among the politicians and voters of the radical right. Politicians from the Sweden Democrats include more outsiders and vulnerable insiders, compared to both the population and, very starkly, other political parties. Over-representation also grows across sub-groups of labor-market outsiders the more they lost (relative to insiders) from the make-work-pay reforms. For voters, we find a strong positive correlation between the Sweden Democrats’ electoral success and the impact of the economic reforms and the financial crisis (i) across municipalities and (ii) across voting districts within municipalities….
One may legitimately ask why new politicians and voters who suffered economic shocks turned to the radical right, rather than the Swedish Left party or the Social Democrats, parties which have traditionally favored redistributive policies and job security (e.g., Guiso et al. 2017). The suggested answer from our analysis is that the political left offers a slate of politicians skewed away from labor market outsiders and vulnerable insiders towards secure insiders….
Our analysis does not show a link between direct exposure to immigration and support for the radical right. The 2002-2014 rise of the Sweden Democrats coincides with a higher tolerance for immigration of the average Swede. The divergent anti-immigration attitudes between insiders and outsiders are also weaker than the divergent trust in government institutions. This does not preclude an instrumental role of anti-immigration rhetoric in the party’s electoral success. With diminished trust in established parties, taking a stand so clearly at odds with the inclusive values of the traditional political class signals an uncompromising opposition to the establishment, which complements the credibility attained by citizen candidates.
Moreover, we could easily think of an indirect link from economic outcomes to attitudes on immigration, and/or the salience of immigration policy in voting decisions. Perhaps most importantly, economic pressures may make people more receptive toward political messages that emphasize the fiscal costs of immigration and the latent redistribution from foreign to native-born by restricting it. Economic shocks could also lead low-income natives away from a social identification with class and support for redistribution, towards an identification with the nation and support for tax cuts and restricted immigration, even at a utility cost (Shayo 2009). Similarly, disgruntled economic losers may socially identify more strongly with their own in-group. If these voters put the blame for their predicament on out-groups, they may be attracted by a party that is anti-establishment as well as anti-immigration. Other explanations along similar lines posit that a long-standing, latent, anti-immigrant stance may interact with economic shocks to drive populist voting (Gidron and Hall 2017).
An equally important, though separate, topic to discuss is the long-term social decline of the white, middle or working-class man, often with short education. This population segment is the backbone of radical right voters, both in Sweden and elsewhere. In the Swedish case, these demographic segments were more likely to suffer from the economic events that we study. We show, however, that the strong relationships of voting and political candidacy remain after flexibly controlling for gender, education, and industrial sector. All in all, our results are compatible with an understanding that both long-term and short-term socioeconomic decline matters, and that short-term economic decline can be a trigger event for so-called “latent” voting segments that have more nostalgia for the past (Norris and Inglehart 2018).
Huffington Post has published an OK short summary of the paper but I strongly recommend diving into the whole thing.
About this website
HUFFINGTONPOST.COM|BY AKBAR SHAHID AHMED
A new study challenges the conventional wisdom that nativist politicians are gaining ground simply because countries like Sweden now have more foreigners.